加载中…
Local spend caps, kill-switch blocking and Ed25519-signed receipts for AI agent tool calls.
AgentGuard 是 ChatHome Skills 收录的 MCP 条目，来源为 Official MCP Registry，质量分 73/100。 作为 Model Context Protocol 服务器，它向 Claude、Cursor、Copilot 等兼容 MCP 的 LLM 客户端暴露可直接调用的工具，让 agent 的能力超出模型内置范围。 该条目由 ChatHome Skills 的每日抓取从官方 registry 和 GitHub 发现查询中收录，每次运行都会重新核验仓库事实、新鲜度信号和安装元数据。 质量分综合了官方 registry 收录情况、可用安装命令、许可证完整性、仓库活跃度和 GitHub star 趋势，完整算法说明见来源页。安装命令已列于上方，可直接复制使用。 在将 AgentGuard 加入生产 agent 工作流前，请核对上方的最近更新日期和仓库状态——下方事实快照是最近一次抓取的时点值。
npx -y @agentguard-run/mcp@0.3.1运行此命令前，请先核对上游源码。
{
"_meta": {
"io.modelcontextprotocol.registry/official": {
"status": "active",
"isLatest": true,
"updatedAt": "2026-09-21T18:24:44.541807Z",
"publishedAt": "2026-09-21T18:24:44.541807Z",
"statusChangedAt": "2026-09-21T18:24:44.541807Z"
}
},
"server": {
"name": "io.github.MerchantGuard/agentguard",
"_meta": {
"io.modelcontextprotocol.registry/publisher-provided": {
"description": "Local spend caps, kill-switch blocking and Ed25519-signed receipts for AI agent tool calls. Runs on the developer's machine. Zero data plane: prompts and tool calls never pass through this server."
}
},
"title": "AgentGuard",
"$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
"version": "0.3.1",
"packages": [
{
"version": "0.3.1",
"transport": {
"type": "stdio"
},
"identifier": "@agentguard-run/mcp",
"runtimeHint": "npx",
"registryType": "npm",
"environmentVariables": [
{
"name": "AGENTGUARD_MCP_DAILY_CAP_CENTS",
"format": "number",
"default": "500",
"description": "Local daily spend cap in cents. A nonnegative integer."
},
{
"name": "AGENTGUARD_MCP_PER_CALL_CAP_CENTS",
"format": "number",
"default": "100",
"description": "Local per-call spend cap in cents. A nonnegative integer."
},
{
"name": "AGENTGUARD_MCP_TENANT",
"format": "string",
"default": "mcp-local",
"description": "Tenant identifier stamped into local receipts."
},
{
"name": "AGENTGUARD_HOME",
"format": "filepath",
"description": "Local signing-key, configuration and ledger directory. Defaults to .agentguard in the operating system home directory."
},
{
"name": "AGENTGUARD_MCP_LEDGER",
"format": "string",
"choices": [
"ndjson",
"memory"
],
"default": "ndjson",
"description": "Local decision storage. ndjson persists receipts across restarts; memory retains this process only."
},
{
"name": "AGENTGUARD_MCP_DISABLE_COST_OVERRIDE",
"format": "string",
"choices": [
"0",
"1"
],
"default": "0",
"description": "Set to 1 to discard and prevent model pricing overrides for this server."
},
{
"name": "AGENTGUARD_LOCK_COST_OVERRIDES",
"format": "string",
"choices": [
"0",
"1"
],
"default": "0",
"description": "Spend SDK startup lock. Set to 1 to prevent model pricing overrides."
},
{
"name": "AGENTGUARD_ACTOR_DIGEST",
"format": "string",
"choices": [
"0",
"1"
],
"default": "0",
"description": "Set to 1 to hash actor identifiers before they enter local receipts."
},
{
"name": "AGENTGUARD_LICENSE_KEY",
"format": "string",
"isSecret": true,
"description": "Optional AgentGuard license key for license validation and seat registration. Otherwise read from local configuration."
},
{
"name": "AGENTGUARD_LICENSE_ENDPOINT",
"format": "string",
"default": "https://agentguard.run",
"description": "License and seat service base URL. No prompts, tool content or receipts are sent."
},
{
"name": "AGENTGUARD_NO_BEACON",
"format": "string",
"choices": [
"0",
"1"
],
"default": "0",
"description": "Set to 1 to disable optional activation telemetry regardless of saved consent."
},
{
"name": "AGENTGUARD_TELEMETRY",
"format": "string",
"choices": [
"0",
"1"
],
"default": "0",
"description": "Set to 1 to opt in to content-free activation telemetry. Otherwise saved consent applies, default off."
},
{
"name": "AGENTGUARD_INSTALL_ID",
"format": "string",
"description": "Optional identifier used only by opted-in activation telemetry when no saved install identifier exists."
},
{
"name": "AGENTGUARD_ANONYMOUS_INSTALL_ID",
"format": "string",
"description": "Legacy fallback for AGENTGUARD_INSTALL_ID in opted-in activation telemetry."
},
{
"name": "CI",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "GITHUB_ACTIONS",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "GITLAB_CI",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "CIRCLECI",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "BUILDKITE",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "VERCEL",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
},
{
"name": "NETLIFY",
"format": "string",
"description": "Existing runtime marker. A nonempty value marks opted-in activation telemetry as CI; it does not enable telemetry."
}
]
}
],
"repository": {
"url": "https://github.com/MerchantGuard/agentguard-mcp",
"source": "github"
},
"websiteUrl": "https://agentguard.run",
"description": "Local spend caps, kill-switch blocking and Ed25519-signed receipts for AI agent tool calls."
}
}